HydroQuantum: A new quantum-driven Python package for hydrological simulation

Abstract

This research aims to leverage the power of quantum computing for hydrological simulation. A new “HydroQuantum” Python package is created to facilitate this implementation, enabling researchers to explore the potential of quantum algorithms in hydrological simulations. “HydroQuantum” was implemented for daily streamflow and stream water temperature (SWT) simulations across continental US. The package includes Variational Quantum Circuits (VQC), a fully quantum Long Short-Term Memory network (QLSTM), and a hybrid quantum-classical LSTM. All algorithms were benchmarked against classical LSTM and trained and tested during 2000–2014 and 2015–2022 for daily streamflow and SWT simulations, respectively. While QLSTM showed impressive results in capturing temporal dependencies in streamflow data, it consistently underperformed classical LSTM for SWT simulation. Sensitivity analysis further revealed that precipitation and snow-water equivalent were two major contributors to quantum-driven simulation. This research explores the potential of quantum computing in complex time series simulations, leading to breakthroughs in hydrological modeling.